Search

Background Glioblastoma multiforme (GBM) tends to occur between the ages of 45 and 70. This relatively early onset and its poor prognosis make the impact of GBM on public health far greater than would be suggested by its relatively low frequency. Tissue and blood samples have now been collected for a number of populations, and predisposing alleles have been sought by several different genome-wide association (GWA) studies. The Cancer Genome Atlas (TCGA) at NIH has also collected a considerable amount of data. Because of the low concordance between the results obtained using different populations, only 14 predisposing single nucleotide polymorphism (SNP) candidates in five genomic regions have been replicated in two or more studies. The purpose of this paper is to present an improved approach to biomarker identification. Methods Association analysis was performed with control of population stratifications using the EIGENSTRAT package, under the null hypothesis of "no association between GBM and control SNP genotypes," based on an additive inheritance model. Genes that are strongly correlated with identified SNPs were determined by linkage disequilibrium (LD) or expression quantitative trait locus (eQTL) analysis. A new approach that combines meta-analysis and pathway enrichment analysis identified additional genes. Results (i) A meta-analysis of SNP data from TCGA and the Adult Glioma Study identifies 12 predisposing SNP candidates, seven of which are reported for the first time. These SNPs fall in five genomic regions (5p15.33, 9p21.3, 1p21.2, 3q26.2 and 7p15.3), three of which have not been previously reported. (ii) 25 genes are strongly correlated with these 12 SNPs, eight of which are known to be cancer-associated. (iii) The relative risk for GBM is highest for risk allele combinations on chromosomes 1 and 9. (iv) A combined meta-analysis/pathway analysis identified an additional four genes. All of these have been identified as cancer-related, but have not been previously associated with glioma. (v) Some SNPs that do not occur reproducibly across populations are in reproducible (invariant) pathways, suggesting that they affect the same biological process, and that population discordance can be partially resolved by evaluating processes rather than genes. Conclusion We have uncovered 29 glioma-associated gene candidates; 12 of them known to be cancer related (p = 1. 4 × 10-6), providing additional statistical support for the relevance of the new candidates. This additional information on risk loci is potentially important for identifying Caucasian individuals at risk for glioma, and for assessing relative risk.

The ability to accurately compute the atomic positions of substrate-bound ligands is central to understanding biological recognition. Although substantial progress has been made in docking small, relatively rigid ligands, the problem of docking flexible peptides remains open. In this communication we present a new method that allows configurational flexibility of peptides, and apply it to predict the conformation of peptides bound to two class-I major histocompatibility complex receptors: human HLA-A2, and murine H-2Kb. Using only the approximate locations of the amino and carboxyl-terminal residues of the bound peptide, our calculations yield structures with backbone conformations that are similar to structures reported crystallographically.

We recently developed a rapid loop closure algorithm in which bond lengths are scaled to constrain the ends of a segment to match a known distance and then gradually relaxed to their standard values, with boundary constraints maintained. Although the algorithm predicted the Zif286 zinc-finger loop to within approximately 2 A, it had a serious limitation that made its more general use tentative: it omitted the atomic environment of the loop. Here we report an extension of the algorithm to take into account the protein environment surrounding a given loop from the outset of the conformational search and show that it predicts structure with an efficiency and accuracy that could not be achieved without continuous environmental inclusion. The algorithm should be widely applicable to structure determination when complete experimental information is unavailable.

We describe, test, and apply a new computational algorithm for generating protein loop conformations subject to distance and secondary structure constraints. The algorithm is based upon initial scaling and subsequent relaxation of covalent bond lengths. The scaling-relaxation procedure needs no additional energy terms and can be readily incorporated into existing molecular modeling packages. The algorithm uses an all-atom energy function from the outset in a straightforward way so that about 60% of the generated loop conformations are free of severe distortions of covalent bond lengths and angles. An extensive application to the major loop conformations of TFIIIA-type zinc fingers (Zif268 and ADR1) is presented, as well as preliminary calculations on hypervariable loops of two immunoglobulins (MCPC603 and Bence-Jones). © 1993 John Wiley & Sons, Inc.

Low energy conformations have been generated for melittin, pancreatic polypeptide, and ribonuclease S-peptide, both in the vicinity of x-ray structures by energy refinement and by an unconstrained search over the entire conformational space. Since the structural polymorphism of these medium-sized peptides in crystal and solution is moderate, comparing the calculated conformations to x-ray and nmr data provides information on local and global behavior of potential functions. Local analysis includes standardization calculations, which show that models with standard geometry can approximate good resolution x-ray data with less than 0.5 A rms deviation (RMSD). However, the atomic coordinates are shifted up to 2 A RMSD by local energy minimization, and thus 2 A is generally the smallest RMSD value one can target in a conformational search using the same energy evaluation models. The unconstrained search was performed by a buildup-type method based on dynamic programming. To accelerate the generation of structures in the conformational search, we used the ECEPP potential, defined in terms of standard polypeptide geometry. A number of low energy conformations were further refined by relaxing the assumption of standard bond lengths and bond angles through the use of the CHARMM potential, and the hydrophobic folding energies of Eisenberg and McLachlan were calculated. Each conformation is described in terms of the RMSD from the native, hydrogen-bonding structure, solvent-acessible surface area, and the ratio of surfaces corresponding to nonpolar and polar residues. The unconstrained search finds conformations that are different from the native, sometimes substantially, and in addition, have lower conformational energies than the native. The origin of deviations is different for each of the three peptides, but in all examples the refined x-ray structures have lower energies than the calculated incorrect folds when (1) the assumption of standard bond lengths and bond angles is relaxed; (2) a small and constant effective dielectric permittivity (e < 10) is used; and (3) the hydrophobic folding energy is incorporated into the potential. © 1993 John Wiley & Sons, Inc.

We develop a general method to identify gene networks from pair-wise correlations between genes in a microarray data set and apply it to a public prostate cancer gene expression data from 69 primary prostate tumors. We define the degree of a node as the number of genes significantly associated with the node and identify hub genes as those with the highest degree. The correlation network was pruned using transcription factor binding information in VisANT (http://visant.bu.edu/) as a biological filter. The reliability of hub genes was determined using a strict permutation test. Separate networks for normal prostate samples, and prostate cancer samples from African Americans (AA) and European Americans (EA) were generated and compared. We found that the same hubs control disease progression in AA and EA networks. Combining AA and EA samples, we generated networks for low low (

We present biological imaging and sensing methods based on optical resonance and interference. In fluorescence microscopy, our nanoscale imaging capability sheds light onto conformational changes of DNA, DNA-protein complexes and polymer coatings on a solid surface. Interference measurements on a layered substrate yield a label-free sensing platform for protein binding in a high-throughput micro-array format.

Energy minimization is one of the main approaches to the computational determination of macromolecular structure. Due to the approximations in the empirical free-energy functions and due to the computational difficulties in locating their global minima, the problem is at present intractable when the only information available is the sequence of subunits forming the molecule. A less-demanding problem in terms of both physics and mathematics is constrained optimization, which uses additional but incomplete experimental information such as distances between certain atoms. This paper reviews methods for generating molecular structure using bond lengths and angles as variables and shows how the structure can be fully specified in terms of local geometry. The analysis permits precise statements to be made about the minimum set of distances that specify a unique structure without recourse to energy minimization. We then discuss the complementary situation, i.e., structure prediction with energy minimization based only on sequence information. Finally, we show how distance constraints can be incorporated into energy minimization methods.

High throughput technologies, including array-based chromatin immunoprecipitation, have rapidly increased our knowledge of transcriptional maps—the identity and location of regulatory binding sites within genomes. Still, the full identification of sites, even in lower eukaryotes, remains largely incomplete. In this paper we develop a supervised learning approach to site identification using support vector machines (SVMs) to combine 26 different data types. A comparison with the standard approach to site identification using position specific scoring matrices (PSSMs) for a set of 104 Saccharomyces cerevisiae regulators indicates that our SVM-based target classification is more sensitive (73 vs. 20%) when specificity and positive predictive value are the same. We have applied our SVM classifier for each transcriptional regulator to all promoters in the yeast genome to obtain thousands of new targets, which are currently being analyzed and refined to limit the risk of classifier over-fitting. For the purpose of illustration we discuss several results, including biochemical pathway predictions for Gcn4 and Rap1. For both transcription factors SVM predictions match well with the known biology of control mechanisms, and possible new roles for these factors are suggested, such as a function for Rap1 in regulating fermentative growth. We also examine the promoter melting temperature curves for the targets of YJR060W, and show that targets of this TF have potentially unique physical properties which distinguish them from other genes. The SVM output automatically provides the means to rank dataset features to identify important biological elements. We use this property to rank classifying k-mers, thereby reconstructing known binding sites for several TFs, and to rank expression experiments, determining the conditions under which Fhl1, the factor responsible for expression of ribosomal protein genes, is active. We can see that targets of Fhl1 are differentially expressed in the chosen conditions as compared to the expression of average and negative set genes. SVM-based classifiers provide a robust framework for analysis of regulatory networks. Processing of classifier outputs can provide high quality predictions and biological insight into functions of particular transcription factors. Future work on this method will focus on increasing the accuracy and quality of predictions using feature reduction and clustering strategies. Since predictions have been made on only 104 TFs in yeast, new classifiers will be built for the remaining 100 factors which have available binding data. Electronic Supplementary Material Supplementary material is available in the online version of this article at http://dx.doi.org/10.1007/s11693-006-9003-3 and is accessible for authorized users.

OBJECTIVES To identify a robust panel of microRNA signatures that can classify tumor from normal kidney using microRNA expression levels. Mounting evidence suggests that microRNAs are key players in essential cellular processes and that their expression pattern can serve as diagnostic biomarkers for cancerous tissues. METHODS We selected 28 clear-cell type human renal cell carcinoma (ccRCC), samples from patient-matched specimens to perform high-throughput, quantitative real-time polymerase chain reaction analysis of microRNA expression levels. The data were subjected to rigorous statistical analyses and hierarchical clustering to produce a discrete set of microRNAs that can robustly distinguish ccRCC from their patient-matched normal kidney tissue samples with high confidence. RESULTS Thirty-five microRNAs were found that can robustly distinguish ccRCC from their patient-matched normal kidney tissue samples with high confidence. Among this set of 35 signature microRNAs, 26 were found to be consistently downregulated and 9 consistently upregulated in ccRCC relative to normal kidney samples. Two microRNAs, namely, MiR-155 and miR-21, commonly found to be upregulated in other cancers, and miR-210, induced by hypoxia, were also identified as overexpressed in ccRCC in our study. MicroRNAs identified as downregulated in our study can be correlated to common chromosome deletions in ccRCC. CONCLUSIONS Our analysis is a comprehensive, statistically relevant study that identifies the microRNAs dysregulated in ccRCC, which can serve as the basis of molecular markers for diagnosis.

We describe a new method based on principal component analysis and robust consensus ensemble clustering to identify and elucidate the subtypes of breast cancer disease. The method was applied to microarray gene expression data using micro-dissection of samples from 36 breast cancer patients with at least two of three pathological stages of disease. Controls were normal breast epithelial cells from 3 disease free patients. Our method identified an optimum set of genes and strong, stable clusters which correlated well with clinical classification into Luminal, Basal and Her2+ subtypes based on ER, PR and Her2 status. It also revealed a hierarchical portrait of disease progression through various grades and stages and identified genes and functional pathways for each stage, grade and disease subtype. We found that gene expression heterogeneity across subtypes is much greater than the heterogeneity of progression from DCIS to IDC within a subtype, suggesting that the disease subtypes are distinct disease processes. The averaging over data perturbations and clustering methods is critical in the robust identification of subtypes and gene markers for grade and progression.

Background Information obtained from diverse data sources can be combined in a principled manner using various machine learning methods to increase the reliability and range of knowledge about protein function. The result is a weighted functional linkage network (FLN) in which linked neighbors share at least one function with high probability. Precision is, however, low. Aiming to provide precise functional annotation for as many proteins as possible, we explore and propose a two-step framework for functional annotation (1) construction of a high-coverage and reliable FLN via machine learning techniques (2) development of a decision rule for the constructed FLN to optimize functional annotation. Results We first apply this framework to Saccharomyces cerevisiae. In the first step, we demonstrate that four commonly used machine learning methods, Linear SVM, Linear Discriminant Analysis, Naïve Bayes, and Neural Network, all combine heterogeneous data to produce reliable and high-coverage FLNs, in which the linkage weight more accurately estimates functional coupling of linked proteins than use individual data sources alone. In the second step, empirical tuning of an adjustable decision rule on the constructed FLN reveals that basing annotation on maximum edge weight results in the most precise annotation at high coverages. In particular at low coverage all rules evaluated perform comparably. At coverage above approximately 50%, however, they diverge rapidly. At full coverage, the maximum weight decision rule still has a precision of approximately 70%, whereas for other methods, precision ranges from a high of slightly more than 30%, down to 3%. In addition, a scoring scheme to estimate the precisions of individual predictions is also provided. Finally, tests of the robustness of the framework indicate that our framework can be successfully applied to less studied organisms. Conclusion We provide a general two-step function-annotation framework, and show that high coverage, high precision annotations can be achieved by constructing a high-coverage and reliable FLN via data integration followed by applying a maximum weight decision rule.

CONTENTS PERSPECTIVES AND OVERVIEW • . . . . . .... 70 The Molecules: a Brief Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Two Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... , . . . . . . . . . . . . . . . . . ,..... ...... . . . . . . . 72 IDENTIFYING IMMUNOGENIC PEPTIDES: RECENT HISTORy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., .... ., .... . . . ., . . , 73 Antigenic Determinants Recognized by Antibodies . Antigenic Determinants Recognized by T Cells ..... . 73 74 THE ANTIGENIC MOTIF: CONCEPTUAL CONSlUEKATIONS . . . . . . . . . . . . . . . . . . . . . . . . . .,., 74 AMPHIPATHICITY: A MAJOR ANTIGENIC MOTIF .. .,... . . . . . . . . . . . . . . . . . . . . . ...... . . . . . . . . . . . . . . . . . . ...... . . . . . . . 75 A GENERAL APPROACH TO IDENTIFYING T CELL ANTIGENIC SITES. . . . . . . . . ... . . . . . . . . . . . . . . . . ......... 76

We identified significantly hypermethylated genes in clear cell renal cell carcinoma.We previously identified a set of under expressed genes in renal cell carcinoma tissue through transcriptional profiling and a robust computational screen. We selected 19 of these genes for hypermethylation analysis using a rigorous search for the best candidate regions, considering CpG islands and transcription factor binding sites. The genes were analyzed for hypermethylation in the DNA of 38 matched clear cell renal cell carcinoma and normal samples using matrix assisted laser desorption ionization time-of-flight mass spectrometry. The significance of hypermethylation was assessed using 3 statistical tests. We validated the down-regulation of significantly hypermethylated genes at the RNA and protein levels in a separate set of patients using reverse transcriptase-polymerase chain reaction, immunohistochemistry and Western blots.We found 7 significantly hypermethylated regions from 6 down-regulated genes, including SFRP1, which was previously shown to be hypermethylated in renal cell carcinoma and other cancer types.To our knowledge we report for the first time that another 5 genes (SCNN1B, SYT6, DACH1, and the tumor suppressors TFAP2A and MT1G) are hypermethylated in renal cell carcinoma. Robust computational screens and the high throughput methylation assay resulted in an enriched set of novel genes that are epigenetically altered in clear cell renal cell carcinoma. Overall the detection of hypermethylation in these highly down-regulated genes suggests that assaying for their methylation using cells from urine or blood could provide the basis for a viable diagnostic test.

Gene expression analysis has identified biologically relevant subclasses of breast cancer. However, most classification schemes do not robustly cluster all HER2+ breast cancers, in part due to limitations and bias of clustering techniques used. In this article, we propose an alternative approach that first separates the HER2+ tumors using a gene amplification signal for Her2/neu amplicon genes and then applies consensus ensemble clustering separately to the HER2+ and HER2− clusters to look for further substructure. We applied this procedure to a microarray data set of 286 early-stage breast cancers treated only with surgery and radiation and identified two basal and four luminal subtypes in the HER2− tumors, as well as two novel and robust HER2+ subtypes. HER2+ subtypes had median distant metastasis-free survival of 99 months [95% confidence interval (95% CI), 83–118 months] and 33 months (95% CI, 11–54 months), respectively, and recurrence rates of 11% and 58%, respectively. The low recurrence subtype had a strong relative overexpression of lymphocyte-associated genes and was also associated with a prominent lymphocytic infiltration on histologic analysis. These data suggest that early-stage HER2+ cancers associated with lymphocytic infiltration are a biologically distinct subtype with an improved natural history. [Cancer Res 2007;67(22):10669–76]

We develop a new technique to analyse microarray data which uses a combination of principal components analysis and consensus ensemble k-clustering to find robust clusters and gene markers in the data. We apply our method to a public microarray breast cancer dataset which has expression levels of genes in normal samples as well as in three pathological stages of disease; namely, atypical ductal hyperplasia or ADH, ductal carcinoma in situ or DCIS and invasive ductal carcinoma or IDC. Our method averages over clustering techniques and data perturbation to find stable, robust clusters and gene markers. We identify the clusters and their pathways with distinct subtypes of breast cancer (Luminal, Basal and Her2+). We confirm that the cancer phenotype develops early (in early hyperplasia or ADH stage) and find from our analysis that each subtype progresses from ADH to DCIS to IDC along its own specific pathway, as if each was a distinct disease.

We introduce a new and potentially valuable method for delineating the repertoire of protein complexes in highly mutable organisms and, in conjunction with other methods, for specifying the structural details of complexes. In the first instance the method provides a guide to selecting proteins for co-crystallization; in the second it augments the collection of structures determined by crystallography and other methods, including the discovery of possible alternative binding sites of known complexes. The key to the method is the availability of multiple sequence variants of an organism-arrived at either naturally or by directed mutagenesis in appropriate laboratory facilities. Amino acids that are important for the structural stability of a protein or complex tend to be conserved, generally mutating only when compensatory changes occur. Consequently significant correlations in variation of two conserved amino acids in the same protein suggest that they interact with one another, either directly or indirectly. Similarly, correlated mutations between conserved amino acids in different proteins suggest that they may be at a site of physical interaction. We have identified all highly conserved 9-11 amino acid long segments from HIV proteins and then identified pairs from different proteins with highly significant co-variation. Using the HIV reverse transcriptase and integrase proteins as an example, we demonstrate how the interface and combining sites can be inferred by co-variation analysis and rigid body docking. The potential significance for antiviral drug and vaccine design is briefly discussed.

Thermodynamic favorability of transcription factor (TF) binding to DNA is a significant factor in the control of gene expression. Theoretical and in vitro measures link the relative equilibrium energy of a particular DNA binding protein to the sequence variation among binding sites in a genome. Extending this principle, we investigate whether biological variation in expression levels of active proteins leads to regulation of different sets of genes, based on inferred affinities of sites upstream of those genes. The TF-concentration-dependent variation in the repertoire of genes regulated by a particular TF is expected to follow patterns of chemical partitioning over DNA sites having differing affinity, and we develop a new modeling approach to test this hypothesis. Based on computational TF binding site discovery and genome-wide expression data available in Saccharomyces cerevisiae, we explore motif content for sets of genes and conditions having varying concentrations of different transcription factors which turn those genes on or off. We find cases of significant correlation between the level of intragenomic motif sequence variation and modeled TF protein levels that actuates regulation of corresponding sets of genes, and discuss the observed TF motif variants for several yeast transcription factors, as well as the potential biological functions of genes that are regulated by differential response to these high and low concentrations of particular TFs. These findings suggest that motif sequences of transcription factor binding sites may often be linked with the expression state of corresponding DNA-binding proteins. Contact: mellor@bu.edu

Phylogenetic profiling is now an effective computational method to detect functional associations between proteins. The method links two proteins in accordance with the similarity of their phyletic distributions across a set of genomes. While pair-wise linkage is useful, it misses correlations in higher order groups: triplets, quadruplets, and so on. Here we assess the probability of observing co-occurrence patterns of 3 binary profiles by chance and show that this probability is asymptotically the same as the mutual information in three profiles. We demonstrate the utility of the probability and the mutual information metrics in detecting overly represented triplets of orthologous proteins which could not be detected using pairwise profiles. These triplets serve as small building blocks, i.e. motifs in protein networks; they allow us to infer the function of uncharacterized members, and facilitate analysis of the local structure and global organization of the protein network. Our method is extendable to N-component clusters, and therefore serves as a general tool for high order protein function annotation.

The common consideration in approaching protection against HIV, whether by a vaccine or therapeutic, is identification of suitable targets. Among the central criteria for suitability is target stability; i.e. resistance to mutation. In this paper we address the problem of stability, and develop methods for identifying stable targets. The targets that we focus on are structures formed by viral peptides and products of the class I major histocompatibility complex, the target of the immune system. The method mines the large databases of fully sequenced HIV genomes and MHC binding peptides, and takes account of human polymorphism to construct hundreds of subpopulation specific stable targets, each consisting of combinations of 3-5 complexes.

Transcription factor binding sites (TFBS) in gene promoter regions are often predicted by using position specific scoring matrices (PSSMs), which summarize sequence patterns of experimentally determined TF binding sites. Although PSSMs are more reliable than simple consensus string matching in predicting a true binding site, they generally result in high numbers of false positive hits. This study attempts to reduce the number of false positive matches and generate new predictions by integrating various types of genomic data by two methods: a Bayesian allocation procedure, and support vector machine classification. Several methods will be explored to strengthen the prediction of a true TFBS in the Saccharomyces cerevisiae genome: binding site degeneracy, binding site conservation, phylogenetic profiling, TF binding site clustering, gene expression profiles, GO functional annotation, and k-mer counts in promoter regions. Binding site degeneracy (or redundancy) refers to the number of times a particular transcription factor's binding motif is discovered in the upstream region of a gene. Phylogenetic conservation takes into account the number of orthologous upstream regions in other genomes that contain a particular binding site. Phylogenetic profiling refers to the presence or absence of a gene across a large set of genomes. Binding site clusters are statistically significant clusters of TF binding sites detected by the algorithm ClusterBuster. Gene expression takes into account the idea that when the gene expression profiles of a transcription factor and a potential target gene are correlated, then it is more likely that the gene is a genuine target. Also, genes with highly correlated expression profiles are often regulated by the same TF(s). The GO annotation data takes advantage of the idea that common transcription targets often have related function. Finally, the distribution of the counts of all k-mers of length 4, 5, and 6 in gene's promoter region were examined as means to predict TF binding. In each case the data are compared to known true positives taken from ChIP-chip data, Transfac, and the Saccharomyces Genome Database. First, degeneracy, conservation, expression, and binding site clusters were examined independently and in combination via Bayesian allocation. Then, binding sites were predicted with a support vector machine (SVM) using all methods alone and in combination. The SVM works best when all genomic data are combined, but can also identify which methods contribute the most to accurate classification. On average, a support vector machine can classify binding sites with high sensitivity and an accuracy of almost 80%.

High-throughput gene expression profiling can identify sets of genes that are differentially expressed between different phenotypes. Discovering marker genes is particularly important in diagnosis of a cancer phenotype. However, gene sets produced to date are too large to be economically viable diagnostics. We use a hybrid decision tree-discriminant analysis to identify small sets of genes, i.e. single genes and gene pairs, which separate normal samples from different stages of tumor samples. Half the samples are selected for training to form the probability distribution of expression values of each gene. The distributions for the tumor and normal phenotypes are then used to classify the test samples. The algorithm also identifies gene pairs by combining the probability distributions to construct a decision tree which is used to determine the class of test samples. After a series of training and testing sessions, genes and gene pairs that classify all samples correctly are recorded. The method was applied to a breast cancer data; and classifier genes that distinguish normal breast from different stages of breast tumor were identified. The genes were ranked according to their minimum Euclidean distance between the expression values in tumor and normal samples. The algorithm was able to pick known cancer related genes but also find genes that were not identified as differentially expressed by t-test with a 2 fold cut-off. Overall, the method generates possible diagnostic genes and gene pairs for a specific disease phenotype to pursue further biological interpretations in cancer biology.

Genome scale identification of transcription factor binding sites (TFBS) is fundamental to understanding the complexities of mRNA expression at both the cell and organismal levels. While high-throughput experimental methods provide associations between transcription factors and the genes they regulate under a specified experimental condition, computational methods are still required to pinpoint the exact location of binding. Moreover, since the binding site is an intrinsic property of the promoter region, computational methods are in principle more general than condition dependent experimental methods. Computational identification of TFBSs is complicated in at least two different ways. First, transcription factors bind a heterogeneous distribution of sites and therefore have a distribution of affinities. Second, the set of sequences for which a common site is to be determined do not all have a site for the TF of interest. In this paper, we evaluate the robustness of TFBS identification with respect to both effects. We show addition of upstream regions that do not have the TFBS destroy the specificity of the predicted binding site. We also propose a method to calculate the distance between position weight matrices that can be used to measure "drift'' from the canonical binding site. The results presented here could be useful in developing future transcription factor binding site identification algorithms.

Problems of inference in systems biology are ideally reduced to formulations which can efficiently represent the features of interest. In the case of predicting gene regulation and pathway networks, an important feature which describes connected genes and proteins is the relationship between active and inactive forms, i.e. between the "on" and "off" states of the components. While not optimal at the limits of resolution, these logical relationships between discrete states can often yield good approximations of the behavior in larger complex systems, where exact representation of measurement relationships may be intractable. We explore techniques for extracting binary state variables from measurement of gene expression, and go on to describe robust measures for statistical significance and information that can be applied to many such types of data. We show how statistical strength and information are equivalent criteria in limiting cases, and demonstrate the application of these measures to simple systems of gene regulation.

ELISA (http://romi.bu.edu/elisa/) is a database that was designed for flexibility in defining interesting queries about protein domain evolution. We have defined and included both the inherent characteristics of the domains such as structure and function and comparisons of these characteristics between domains. Thus, the database is useful in defining structural and functional links between related protein domains and by extension sequences that encode them. In this database we introduce and employ a novel method of functional annotation and comparison. For each protein domain we create a probabilistic functional annotation tree using GO. We have designed an algorithm that accurately compares these trees and thus provides a measure of "functional distance" between two protein domains. Along with functional annotation, we have also included structural comparison between protein domains and best sequence comparisons to all known genomes. The latter enables researchers to dynamically do searches for domains sharing similar phylogenetic profiles. This combination of data and tools enables the researcher to design complex queries to carry out research in the areas of protein domain evolution, structure prediction and functional annotation of novel sequences.

A question of fundamental importance is the definition and identification of modules from microarray experiments. A wide variety of techniques have been used to gain insight into the elucidation of such modules. One problem, however, is the inability to directly compare results between the different data sets produced due to the inherent parameterizations of their approaches. We first aim to provide a mechanism by which different approaches to module finding can be directly compared. Moreover, the same approach can be used to internally compare the modules predicted by the same technique, but at different parameterizations. We apply this approach to analyze the flow of genes through modules at different module thresholds of the Barkai Signature method, thereby further resolving the modules into sets of co-expressed genes.

The paper describes the resonant cavity imaging biosensor (RCIB) for binding site detection. The RCIB achieves excellent sensitivity with the use of an optical cavity. Light resonating inside the optical cavity interacts with the intracavity medium, the biomolecules in this case, over the number of passes proportional to the cavity finesse

Motivation: Genes with identical patterns of occurrence across the phyla tend to function together in the same protein complexes or participate in the same biochemical pathway. However, the requirement that the profiles be identical (i) severely restricts the number of functional links that can be established by such phylogenetic profiling; (ii) limits detection to very strong functional links, failing to capture relations between genes that are not in the same pathway, but nevertheless subserve a common function and (iii) misses relations between analogous genes. Here we present and apply a method for relaxing the restriction, based on the probability that a given arbitrary degree of similarity between two profiles would occur by chance, with no biological pressure. Function is then inferred at any desired level of confidence. Results: We derive an expression for the probability distribution of a given number of chance co-occurrences of a pair of non-homologous orthologs across a set of genomes. The method is applied to 2905 clusters of orthologous genes (COGs) from 44 fully sequenced microbial genomes representing all three domains of life. Among the results are the following. (1) Of the 51 000 annotated intrapathway gene pairs, 8935 are linked at a level of significance of 0.01. This is over 30-fold greater than the 271 intrapathway pairs obtained at the same confidence level when identical profiles are used. (2) Of the 540 000 interpathway genes pairs, some 65 000 are linked at the 0.01 level of significance, some 12 standard deviations beyond the number expected by chance at this confidence level. We speculate that many of these links involve nearest-neighbor path, and discuss some examples. (3) The difference in the percentage of linked interpathway and intrapathway genes is highly significant, consistent with the intuitive expectation that genes in the same pathway are generally under greater selective pressure than those that are not. (4) The method appears to recover well metabolic networks. This is illustrated by the TCA cycle which is recovered as a highly connected, weighted edge network of 30 of its 31 COGs. (5) The fraction of pairs having a common pathway is a symmetric function of the Hamming distance between their profiles. This finding, that the functional correlation between profiles with near maximum Hamming distance is as large as between profiles with near zero Hamming distance, and as statistically significant, is plausibly explained if the former group represents analogous genes. Contact: delisi@bu.edu

We present an analysis of physical chemical constraints on the accuracy of DNA micro-arrays under equilibrium and nonequilibrium conditions. At the beginning of the article we describe an algorithm for choosing a probe set with high specificity for targeted genes under equilibrium conditions. The algorithm as well as existing methods is used to select probes from the full Saccharomyces cerevisiae genome, and these probe sets, along with a randomly selected set, are used to simulate array experiments and identify sources of error. Inasmuch as specificity and sensitivity are maximum at thermodynamic equilibrium, we are particularly interested in the factors that affect the approach to equilibrium. These are analyzed later in the article, where we develop and apply a rapidly executable method to simulate the kinetics of hybridization on a solid phase support. Although the difference between solution phase and solid phase hybridization is of little consequence for specificity and sensitivity when equilibrium is achieved, the kinetics of hybridization has a pronounced effect on both. We first use the model to estimate the effects of diffusion, crosshybridization, relaxation time, and target concentration on the hybridization kinetics, and then investigate the effects of the most important kinetic parameters on specificity. We find even when using probe sets that have high specificity at equilibrium that substantial crosshybridization is present under nonequilibrium conditions. Although those complexes that differ from perfect complementarity by more than a single base do not contribute to sources of error at equilibrium, they slow the approach to equilibrium dramatically and confound interpretation of the data when they dissociate on a time scale comparable to the time of the experiment. For the best probe set, our simulation shows that steady-state behavior is obtained in a relaxation time of approximately 12-15 h for experimental target concentrations approximately (10(-13) - 10(-14))M, but the time is greater for lower target concentrations in the range (10(-15)-10(-16))M. The result points to an asymmetry in the accuracy with which up- and downregulated genes are identified.

High-throughput genomic technologies are revolutionizing modern biology. In particular, DNA microarrays have become one of the most powerful tools for profiling global mRNA expression in different tissues and environmental conditions, and for detecting single nucleotide polymorphisms. The broad applicability of gene expression profiling to the biological and medical realms has generated expanding demand for mass production of microarrays, which in turn has created considerable interest in improving the cost effectiveness of microarray fabrication techniques. We have developed the computational framework for an optimal synthesis strategy for oligonucleotide microarrays. The problem was introduced by Hubbell et al. Here, we formalize the problem, obtain precise bounds on its complexity and devise several computational solutions.

Conservation of proximity of a pair of genes across multiple genomes generally indicates that their functions could be linked. Here, we present a systematic evaluation using 42 complete microbial genomes from 25 phylogenetic groups to test the reliability of this observation in predicting function for genes. We find a relationship between the number of phylogenetic groups in which a gene pair is proximate and the probability that the pair belongs to a common pathway. Our method produces 1586 links between ortholog families substantiated by observed proximity in genomes representing at least three phylogenetic groups. Of the pairs annotated in the KEGG database, 80% are in the same biological pathway in KEGG.

Comparative genomics provides at least three methods for identifying functional links between genes: examination of phylogenetic distributions, analysis of conserved proximity and observations of fusions of genes into a multidomain gene in another organism. We show that the functional networks obtained by applying these methods have different topologies and that the information they provide is largely additive. In particular, the combined networks of functional links contain an average of 57% of an organism's complete genetic complement, uncover substantial portions of known pathways, and suggest the function of previously unannotated genes. In addition, the combined networks are qualitatively different from the networks obtained using individual methods., Background Comparative genomics provides at least three methods beyond traditional sequence similarity for identifying functional links between genes: the examination of common phylogenetic distributions, the analysis of conserved proximity along the chromosomes of multiple genomes, and observations of fusions of genes into a multidomain gene in another organism. We have previously generated the links according to each of these methods individually for 43 known microbial genomes. Here we combine these results to construct networks of functional associations. Results We show that the functional networks obtained by applying these methods have different topologies and that the information they provide is largely additive. In particular, the combined networks of functional links contain an average of 57% of an organism's complete genetic complement, uncover substantial portions of known pathways, and suggest the function of previously unannotated genes. In addition, the combined networks are qualitatively different from the networks obtained using individual methods. They have a dominant cluster that contains approximately 80%-90% of the genes, independent of genome size, and the dominant clusters show the small world behavior expected of a biological system, with global connectivity that is nearly random, and local properties that are highly ordered. Conclusions When the information on functional linkage provided by three emerging computational methods is combined, the integrated network uncovers large numbers of conserved pathways and identifies clusters of functionally related genes. It therefore shows considerable utility and promise as a tool for understanding genomic structure, and for guiding high throughput experimental investigations.

Recent advances in massively parallel experimental and computational technologies are leading to radically new approaches to the early phases of the drug production pipeline. The revolution in DNA microarray technologies and the imminent emergence of its analogue for proteins, along with machine learning algorithms, promise rapid acceleration in the identification of potential drug targets, and in high-throughput screens for subpopulation-specific toxicity. Similarly, advances in structural genomics in conjunction with in vitro and in silico evolutionary methods will rapidly accelerate the number of lead drug candidates and substantially augment their target specificity. Taken collectively, these advances will usher in an era of predictive medicine, which will move medical practice from reactive therapy after disease onset, to proactive prevention.

Recent work in computational genomics has shown that a functional association between two genes can be derived from the existence of a fusion of the two as one continuous sequence in another genome. For each of 30 completely sequenced microbial genomes, we established all such fusion links among its genes and determined the distribution of links within and among 15 broad functional categories. We found that 72% of all fusion links related genes of the same functional category. A comparison of the distribution of links to simulations on the basis of a random model further confirmed the significance of intracategory fusion links. Where a gene of annotated function is linked to an unclassified gene, the fusion link suggests that the two genes belong to the same functional category. The predictions based on fusion links are shown here for Methanobacterium thermoautotrophicum , and another 661 predictions are available at http://fusion.bu.edu .

The ability of a protein to function in its biochemical role(s) is determined by its geometry, which in turn is determined by the amino acid sequence and the environment. Amino acid mutations affect the thermodynamics of folding and their effects can be investigated experimentally by such techniques as site-specific mutation, random point mutations and shuffling. Keywords: protein stability; folding free energy; solvation; electrostatic interactions; site-specific mutagenesis; random mutations

Advances in methods of structure determination have led to the accumulation of large amounts of protein structural data. Some 500 distinct protein folds have now been characterized, representing one-third of all globular folds that exist. The range of known structural types and the relatively large fraction of the protein universe that has already been sampled have greatly facilitated the discovery of some unifying principles governing protein structure and evolutionary relationships. These include a highly skewed distribution of topological arrangements of secondary-structure elements that favors a few very common connectivities and a highly skewed distribution in the capacity of folds to accommodate unrelated sequences. These and other observations suggest that the number of folds is far fewer than the number of genes, and that the fold universe is dominated by a small number of giant attractors that accommodate large numbers of unrelated sequences. Thus all basic protein folds will likely be determined in the near future, laying the foundation for a comprehensive understanding of the biochemical and cellular functions of whole organisms.

We present a rapidly executable minimal binding energy model for molecular docking and use it to explore the energy landscape in the vicinity of the binding sites of four different enzyme inhibitor complexes. The structures of the complexes are calculated starting with the crystal structures of the free monomers, using DOCK 4.0 to generate a large number of potential configurations, and screening with the binding energy target function. In order to investigate possible correlations between energy and variation from the native structure, we introduce a new measure of similarity, which removes many of the difficulties associated with root mean square deviation. The analysis uncovers energy gradients, or funnels, near the binding site, with decreasing energy as the degree of similarity between the native and docked structures increases. Such energy funnels can increase the number of random collisions that may evolve into productive stable complex, and indicate that short-range interactions in the precomplexes can contribute to the association rate. The finding could provide an explanation for the relatively rapid association rates that are observed even in the absence of long-range electrostatic steering.

The 1980s saw plenty of discussion on sequencing the human genome. But, according to Charles DeLisi, one conference was crucial for converting an idea to reality.

The peptides that bind class I MHC molecules are restricted in length and often contain key amino acids, anchor residues, at particular positions. The side-chains of peptide anchor residues interact with the polymorphic complementary pockets in MHC peptide-binding grooves and provide the molecular basis for allele-specific recognition of antigenic peptides. We establish correlations between class I MHC specificities for anchor residues and class I MHC sequence markers that occur at the polymorphic positions lining the structural pockets. By analyzing the pocket structures of nine crystallized class I MHC molecules and the modeled structures of another 39 class I MHC molecules, we show that class I pockets can be classified into families that are distinguishable by their common physico-chemical properties and peptide side-chain selectivities. The identification of recurrent structural principles among class I pockets makes it possible to greatly expand the repertoire of known peptide-binding motifs of class I MHC molecules. The evolutionary strategies underlying the emergence of pocket families is briefly discussed.

An effective free energy potential, developed originally for binding free energy calculation, is compared to calorimetric data on protein unfolding, described by a linear combination of changes in polar and nonpolar surface areas. The potential consists of a molecular mechanics energy term calculated for a reference medium (vapor or nonpolar liquid), and empirical terms representing solvation and entropic effects. It is shown that, under suitable conditions, the free energy function agrees well with the calorimetric expression. An additional result of the comparison is an independent estimate of the side-chain entropy loss, which is shown to agree with a structure-based entropy scale. These findings confirm that simple functions can be used to estimate the free energy change in complex systems, and that a binding free energy evaluation model can describe the thermodynamics of protein unfolding correctly. Furthermore, it is shown that folding and binding leave the sum of solute-solute and solute-solvent van der Waals interactions nearly invariant and, due to this invariance, it may be advantageous to use a nonpolar liquid rather than vacuum as the reference medium.

Knowledge of human leucocyte antigen (HLA) peptide binding motifs permits rapid selection of candidate viral protein fragments for induction of T cell-mediated immunity. A search for HLA class I peptide binding motifs in structural proteins of human immunodeficiency virus (HIV) of different genetic lineages provides a map of the genetic organization of potential T cell antigenic sites, and at the same time identifies all motifs in highly conserved regions of HIV-1 env, gag and pol. The density of motifs is anomalous at both the high and low end of the spectrum: local organization is characterized by clustering in relatively short regions, while large scale organization is characterized by anomalously long runs between motifs. The former is expected simply due to the fact that motifs often have overlapping anchor residue sets. A detailed statistical analysis of the latter, however, shows that the length of the runs cannot be accounted for by chance alone. Although motif clusters show no preference to be in either conserved or variable regions, low motif density stretches occur preferentially in variable portions of the protein sequence, which suggests that the virus may be mutating to evade the cellular arm of the immune system.

We report a new free energy decomposition that includes structure-derived atomic contact energies for the desolvation component, and show that it applies equally well to the analysis of single-domain protein folding and to the binding of flexible peptides to proteins. Specifically, we selected the 17 single-domain proteins for which the three-dimensional structures and thermodynamic unfolding free energies are available. By calculating all terms except the backbone conformational entropy change and comparing the result to the experimentally measured free energy, we estimated that the mean entropy gain by the backbone chain upon unfolding (delta Sbb) is 5.3 cal/K per mole of residue, and that the average backbone entropy for glycine is 6.7 cal/K. Both numbers are in close agreement with recent estimates made by entirely different methods, suggesting a promising degree of consistency between data obtained from disparate sources. In addition, a quantitative analysis of the folding free energy indicates that the unfavorable backbone entropy for each of the proteins is balanced predominantly by favorable backbone interactions. Finally, because the binding of flexible peptides to receptors is physically similar to folding, the free energy function should, in principle, be equally applicable to flexible docking. By combining atomic contact energies, electrostatics, and sequence-dependent backbone entropy, we calculated a priori the free energy changes associated with the binding of four different peptides to HLA-A2, 1 MHC molecule and found agreement with experiment to within 10% without parameter adjustment.

Peptides that bind to major histocompatibility complex products (MHC) are known to exhibit certain sequence motifs which, though common, are neither necessary nor sufficient for binding: MHCs bind certain peptides that do not have the characteristic motifs and only about 30% of the peptides having the required motif, bind. In order to develop and test more accurate methods we measured the binding affinity of 463 nonamer peptides to HLA-A2.1. We describe two methods for predicting whether a given peptide will bind to an MHC and apply them to these peptides. One method is based on simulating a neural network and another, called the polynomial method, is based on statistical parameter estimation assuming independent binding of the side-chains of residues. We compare these methods with each other and with standard motif-based methods. The two methods are complementary, and both are superior to sequence motifs. The neural net is superior to simple motif searches in eliminating false positives. Its behavior can be coarsely tuned to the strength of binding desired and it is extendable in a straightforward fashion to other alleles. The polynomial method, on the other hand, has high sensitivity and is a superior method for eliminating false negatives. We discuss the validity of the independent binding assumption in such predictions.

We show that a rapidly executable computational procedure provides the basis for a predictive understanding of antigenic peptide side chain specificity, for binding to class I major histocompatibility complex (MHC) molecules. The procedure consists of a combined search to identify the joint conformations of peptide side chains and side chains comprising the MHC pocket, followed by conformational selection, using a target function, based on solvation energies and modified electrostatic energies. The method was applied to the B pocket region of five MHC molecules, which were chosen to encompass the full range of specificities displayed by anchors at peptide position 2. These were a medium hydrophobic residue (Leu or Met) for HLA-A*0201, a basic residue (Arg or Lys) for HLA-B*2705; a small hydrophobic residue (Val) for HLA-A*6801, an acidic residue (Glu) for HLA-B*4001 and a bulky residue (Tyr) for H-2K(d). The observed anchors are correctly predicted in each case. The agreement for HLA-B40 and H-2K(d) is especially promising, since their structures have not yet been determined experimentally. Because the experimental determination of motifs by elution is difficult and these calculations take only hours on a high speed workstation, the results open the possibility of routine determination of motifs computationally.

We have developed a method that utilizes site-specific mutation data, sequence analysis, immunological data and free-energy minimization, to determine structural features of the ternary complex formed by the T-cell receptor (TcR) and the class I major histocompatibility complex (MHC) molecule bound by peptide. The analysis focuses on the mouse Kd MHC system, for which a large set of clones with sequenced T-cell receptors is available for specific peptides. The general philosophy is to reduce the uncertainties and computation time in a free-energy minimization procedure by identifying and imposing experimental constraints. In addition to assessing compatibility with various kinds of immunological data, we are particularly interested in differentiating the structural features peculiar to this particular system from generic features, and in ascertaining the robustness of the structure; i.e. determining, in so far as possible, the variations in the structure that leave its compatibility with experiment unaltered from those that do not. This last is equivalent to recognizing that certain features of the model are presented with a reasonable degree of confidence, while others remain highly tentative. The central conclusion in the former category is a placement of the TcR on the Kd peptide complex, which has its beta 2, beta 3 and alpha 3 loops (i.e. the second and third complementarity-determining region of the TcR beta chain, and the third complementarity-determining region of the alpha chain) covering the peptide; the alpha 1 and alpha 2 loops covering the MHC alpha 1 helix; the alpha 2 loop interacting with residues on the MHC beta sheet; and the beta 1 and (part of) the beta 2 loops covering the alpha 2 MHC helix. More specifically, our findings include the following. (1) A highly conserved histidine residue in the first complementarity-determining region of the TcR beta chain (beta:CDR1) points outward and interacts with highly conserved side-chains on the MHC alpha 2 helix. (2) The amino-terminal portion of the beta 2 loop interacts with the carboxyl portion of the peptide. A particularly important interaction is K4 of the loop interacting with E8 of the peptide. (3) Charged side-chains of the 11-residue TcR alpha 2 loop interact with conserved charged side-chains at positions 44, 58, 61 and 68 on the MHC. (4) The TcR beta 3 loop interacts with the amino-terminal part of the peptide, up through position 4. (5) the TcR alpha 3 loop interacts with the central portion of the peptide and stacks against the beta 2 loop. (6) Because of the interaction between the beta 2 loop and the peptide, and stacking of beta 2 on alpha 3, alpha 3 gene and V beta gene selection can be correlated. (7) Using the topology of the recently solved TcR alpha chain we predict that the alpha 2 loop interacts with the loop on the MHC beta sheet floor, which encompasses residues 42 to 44.

Free energy maps of the binding site are constructed for class I major histocompatibility complex (MHC) proteins, by rotating and translating amino acid probes along the cleft, and performing a side-chain conformational search at each position. The free energy maps are used to determine favorable residue positions that are then combined to form docked peptide conformations. Because the generic backbone structural motif of peptides bound to class I MHC is known, the mapping is restricted to appropriate regions of the site, but allows for the sometimes substantial variations in backbone and side-chain conformations. In a test demonstrating the quality of predictions for a known MHC site using only a rotational and conformational search, we started from the crystal structure of the HIV-1 gp120/HLA-A2 complex, and predicted the HLA-A2 bound structures of peptides from the influenza matrix protein, the HIV-1 reverse transcriptase, and the human T cell leukemia virus. The calculated peptides are at 1.6, 1.3, and 1.4 A all-atom RMSDs from their respective crystal structures (Madden DR, Garboczi DN, Wiley DC, 1993). A further test, which also included a local translational search, predicted structures across MHCs. In particular, we obtained the Kb/SEV-9 complex (Fremont DH et al., 1992, Science 257:919-927) starting with the complex between HLA-B27 and a generic peptide (Madden DR, Gorga JC, Strominger JL, Wiley DC, 1991, Nature (Lond) 353:321-325), with an all-atom RMSD of 1.2 A, indicating that the docking procedure is essentially as effective for predictions across MHCs as it is for determinations within the same MHC, although at substantially greater computational cost. The requirements for further improvement in accuracy are identified and discussed briefly.